Operant Reinforcement And Respondentswith Similar Topographies

The newborn infant's crying, as its eating, is reflexive, elicited by loud sounds, trauma, extremes of temperature, or food deprivation. Later, however, crying comes under the control of operant as well as respondent reinforcement, because it changes the child's environment through the mediation of an adult. The distinction between crying as an operant or a reflex response is made by noticing whether the crying is a change in the child's behavior caused by (in response to) the external environment or whether it is maintained because it, in itself, produces an effect on the external environment which, in turn, increases its frequency. Very early in the child's life crying ceases to become solely reflexive in nature and occurs because of the specific consequences which follow it. The child who has not eaten in several hours cries both because in the past such crying has led to food and because of the elicitation of the direct effects of food deprivation. As with sucking, there is the possibility of differential reinforcement of a more intense and finely differentiated crying pattern. It is no accident that the cry of the infant is especially aversive to the child's own parent. Those variations in the quality and intensity of the child's crying which are most aversive to the parent are most likely to be followed by the parent's attention and such consequences as food or a dry diaper. Conversely, minimal forms are likely to go unreinforced and will tend to disappear. The resulting crying is idiosyncratic, differentially reinforced by the parent who reacts to those forms of crying which are especially aversive.

A given response such as crying may be simultaneously elicited as in a reflex and reinforced as in an operant response, but in many cases it may not be clear as to the source from which the behavior derives its major strength. The infant who is crying because it has not eaten in several hours is emitting a response which could be elicited solely by the state of its gastrointestinal system. At the same time, however, food deprivation increases the frequency of all those operant responses which in the past have led to eating. We do not know what the exact relationship is between the response topographies in elicited crying, as compared with those in operant crying, or the extent to which they interact. Whether or not differences occur, however, the two topographies must be very similar, because operant crying will inevitably mimic the emotional, "natural" form of crying since the latter has the largest effect on the parent.

To determine whether or not a response is operant or reflex depends ultimately on finding out how to change its frequency and magnitude. To prove that a response is maintained by operant reinforcement rather than as a reflex, one must identify some immediate consequence of the response of the environment, interrupt it, and record a declining frequency of occurrence of the response as a result of its subsequent nonreinforcement. A second method would be to show a progressive shift in topography as the environmental change is made contingent on corresponding progressive changes in response topography. In the reflex, the form of the behavior will be relatively fixed, but its magnitude will increase with increases in the eliciting condition. In actual practice, testing the nature of a response by extinction is very difficult because many schedules of operant reinforcement generate very large dispositions to engage in the behavior which persists for many thousands of responses after reinforcement is no longer continued. If extinction were not carried out long enough, it would not be clear whether the response was being elicited or was being strengthened because extinction had not had its ultimate effect.

Summary Of The Difference Betweenoperant And Respondent Behavior

ment. Operant behavior actually operates on and changes the environment, while reflex behavior represents an effect of the environment on the organism. Moreover, the operant and respondent repertoires represent very different behavioral processes in respect to the specific ways the behaviors are related to the environment.

The form of the reflex is determined by the eliciting stimulus and has its basis in the phylogenetic history of the organism. The reflex is elicited by its stimulus in the sense that there is a one-to-one correspondence between a stimulus and a response. In general, given the stimulus, we know what response will be elicited. Given the response, we know that only a limited number of stimuli will elicit it. The conditioned reflex represents a pairing of stimuli, so that a new stimulus comes to elicit the reflex in a manner similar to that of the unconditioned stimulus. Operant behavior, on the contrary, is maintained by its effect on the environment, rather than through its being elicited by a stimulus. A stimulus following a response increases the subsequent frequency.

The major indication of effect of conditioning in the case of reflex or respondent behavior is the order of magnitude of the unconditioned response of the reflex or the latency of the reflex. In operant behavior, the major effect of reinforcement or conditioning is on the subsequent frequency of occurrence of the behavior. Very strong behavior may have either a small or a large order of magnitude, for example, the neurophysiologist who devotes his life to work with microelectrodes or the surgeon who spends years on technics for suturing small arteries. We evaluate a man's disposition to play golf or bridge by observing how frequently he plays.

The relation to prior stimuli is markedly different for the two kinds of behavior. The magnitude of the reflex is a function of the magnitude of the stimulus. The deflection of the leg increases with the force of the tap on the patellar tendon. In operant behavior, however, very slight stimuli may have major effects, and the relation between magnitude of stimulus and order of magnitude of effect is an arbitrary one, depending on the conditions of reinforcement.

Some performances are dually controlled by both operant and respondent processes, as, for example, the eye blink. Little is known about the interaction of the two processes in these cases.

While we can analyze the properties of operant and respondent behavior separately, the two are generally operating simultaneously in the complex case and it is only by experimental analysis that their separate properties may be discovered. Consider, for example, the experiment of Pavlov demonstrating the conditioning of the reflex action of the parotid gland when food is placed in the dog's mouth. Salivation is elicited and any stimuli accompanying the presentation of food in the mouth soon comes to acquire control over the salivatory response. At the same time, however, any operant activity of the organism (striated muscle) at the time the food is placed in the mouth will be conditioned via operant reinforcement. If the dog happens to turn its head at the instant food is placed in its mouth, the subsequent frequency of head turning is likely to increase. In the converse case, the pigeon whose neck-stretching delivered food would also be subject to the conditions of the development of a conditioned salivatory reflex. Any consistent relation between stimuli present when the neck-stretching occurred and the occurrence of food in the mouth would also produce salivation.

13